Identification of vocal nodules and laryngitis by Gauss mixture model
Tianqi Xu, Keqiao Feng, Yiwen Ge, Xiaojun Zhang, Zhi Tao · 2017
In this paper, Gauss mixture model algorithm is used for recognition of vocal nodules and laryngitis by calculating the posterior probability of training model. The pathological voices database of Soochow University was used in experiment. The experiment result shows that compared with three kinds of common traditional recognition algorithm C4.5 decision trees, Naïve Bayes and Support vector machine. This method to identify the sample recognition rate was increased by 3%, 8% and 9%, which indicated that the Gauss mixture model algorithm is more suitable for the recognition of vocal nodules and laryngitis in this paper, Gauss mixture model algorithm is used for recognition of vocal nodules and laryngitis by calculating the posterior probability of training model. The pathological voices database of Soochow University was used in experiment. The experiment result shows that compared with three kinds of common traditional recognition algorithm C4.5 decision trees, Naïve Bayes and Support vector machine. This method to identify the sample recognition rate was increased by 3%, 8% and 9%, which indicated that the Gauss mixture model algorithm is more suitable for the recognition of vocal nodules and laryngitis.